We have an ongoing program to generate human antibodies to HlV-1 from phage display libraries. The antibodies are intended for clinical use and as tools to explore scientific questions related to HIV-1 infection. The clinical use includes passive immunization to reduce viral load in infected persons and prophylaxis in cases of accidental exposure to virus and in pregnant women to prevent transmission of the virus to offspring. The scientific questions to be explored include defining the epitopes on HIV-1 envelope recognized by human antibodies and understanding why some antibodies are particularly potent in neutralizing virus. These questions pertain directly to vaccine design. The library approach has already generated an antibody capable of broad neutralization of a range of primary isolates of HIV-1 in vitro and protection against infection by a laboratory strain of HIV-1 in the HuSCID mouse model. The antibody has now been improved in terms of both affinity and potency by a procedure of in vitro evolution. Recent data on the brisk production of virus in HIV-1 infection have highlighted the potential value of reducing viral load in limiting disease. Rapid turnover however also increases the problems of neutralization escape so that a single antibody, even if potent, is likely to have limited beneficial effect. Our strategy is to use phage display libraries to isolate a number of antibodies against distinct envelope epitopes with the ability to potently neutralize primary isolates of HIV-1. It is envisaged that a cocktail of these antibodies, or in vitro evolved versions, will be used in combination with other anti-retroviral agents as prophylactic and therapeutic agents in HIV-1 infection. The first specific aim of the proposal is to evaluate a number of promising antibodies (Fabs) generated in the first phase of the project particularly in terms of neutralization of primary isolates, neutralization escape and potential mimetopes. The second aim is to generate new antibodies, particularly from long-term non-progressors, for evaluation. The third aim is to evaluate in vitro evolved and whole antibody versions of the Fabs from the first two aims for inclusion in a cocktail of antibodies for therapeutic and prophylactic use. The fourth aim is to explore features of antibody recognition of viral envelope which lead to potent neutralization of primary isolates. An understanding of this relationship may eventually lead to small molecules capable of mimicking the effect of antibody.